The objective of this project is to determine whether the integration of quantitative characteristics derived rom neuroimaging parameters (magnetic resonance spectroscopic indices, cerebral blood volume and apparent diffusion coefficient), with tissue biomarkers and are predictive of the biologic behavior for patients with glioblastoma multiforme (GBM) post therapy. Standard magnetic resonance imaging (MRI) is limited in assessing the biological behavior of increased enhancement which universally occurs post therapy. Although in the majority of cases this is secondary to tumor progression, following focal therapies such as convection enhanced delivery(CED) of agents, increased enhancement could be secondary to treatment effects. Pathologic specimens obtained post therapy in areas of new enhancement often consist of both 'tumor" and " treatment effect". There are currently no well-established biomarkers that distinguish between tumor or treatment effect or are predictive of the behavior of the radiographic changes post therapy. In this project the objectives will be accomplished by the evaluation of metabolic and physiologic imaging parameters to differentiate between treatment related changes and recurrent tumor. If physiologic imaging can be correlated with tumor biology or clinical outcome, this will allow for the repeated use of a safe, non- invasive, easily acquired biomarker. This will have significant implications to the care of patients and the design, conduct and interpretation of clinical trials in Neuro-Oncology. Patients with radiographically defined progression" will be studied prior to treatment for recurrence and followed for clinical outcome to determine if specific parameters are predictive of biological behavior. This information will give insight into the imaging methods required for evaluation of the effects of CED of a novel compound and prospective physiologic imaging is planned before and after treatment in a clinical trial of CED. These studies will require the support of the Clinical Services Core and the Imaging and Tissue Correlate Core